XyDromatics SR Engine —
structured reports, specialized.
The structured-report specialist. Receive DICOM SR from
modalities, AI vendors, voice-recognition platforms, and
reporting systems; transform with template awareness; fan
out to PACS, VNA, dose-tracking platforms, EMR, and the
ACR Dose Index Registry. A standalone licensable subset of
the Router optimized for SR-specific workflows.
Same platform features as the rest of the router-pattern
family — keyed-hash pseudonymization, built-in
honest-broker capability, shared admin shell, family-wide
site encryption key for cross-product anon-ID consistency.
What SR Engine adds beyond Router’s general DICOM
routing: template awareness, dose-SR specialization, CAD-
finding routing, and the AI-finding round-trip pattern. All
of it with the same platform features as the rest of the
router-pattern family.
SR ingest
Receive DICOM SR objects from any conformant producer — modalities, voice-recognition platforms, AI applications, reporting platforms.
C-STORE SCP for DICOM SR (full SOP class coverage)
STOW-RS for HTTP-based SR submission
Per-AE-Title access control with configurable allowlists
SR template recognition (TID 1500, TID 2000, TID 5200, TID 10001 dose, etc.)
Backpressure semantics during burst-load periods
Non-SR objects rejected at ingest with clear protocol-level errors
Template-aware transformation
SR-specific transformations that respect template structure rather than treating SR as opaque DICOM. Required for cross-vendor SR cleanup and template upgrades.
Template-version migration (e.g., older TID 1500 → current TID 1500)
Cross-vendor SR normalization (different vendors emit subtly different SR for the same intent)
Tag-level rewrite within SR content sequences
Anonymization of patient-identifying SR content with keyed-hash pseudonymization
Per-template transformation rules
Audit log of every transformation applied to SR content
Multi-consumer forwarding + format translation
A single SR object often needs to land in multiple downstream systems — PACS, VNA, dose-tracking platforms, EMR, reporting / dictation platforms. SR Engine fans out without the producer having to, and translates SR into HL7 or FHIR for consumers that don't take DICOM SR natively.
Transform SR → HL7 ORU and emit HL7 to systems that want HL7 messages instead of SR objects
Transform SR → FHIR DiagnosticReport / Observation for FHIR-aware downstreams
Per-destination shape adjustments (different consumers want the same content in different formats)
Failed-destination retry with dead-letter queue + throughput throttling per destination
Confirmation ACK back to producer where required
Radiation-dose SR routing — multiple destination options
Radiation Dose SR (RDSR, TID 10001) is the highest-volume SR type in many imaging departments. SR Engine handles the routing; the customer's ACR / state requirements may be satisfiable in several different ways depending on what other systems they have.
RDSR template recognition with per-modality variants
Forward to dose-tracking platforms (Bayer Radimetrics, Sectra DoseTrack, Siemens teamplay) — for full dose-abnormality monitoring and longitudinal tracking
Forward to EMR (Epic Radiant has integrated dose tracking with ACR DIR submission, automatic RDSR capture, and XR-29 / XR-25 Dose Check alerts) — for institutions on Epic Radiant, this can eliminate the need for a separate dose-tracking platform entirely
Direct encrypted submission to ACR Dose Index Registry — for institutions whose only requirement is the ACR DIR submission itself, no separate dose-tracking platform required
Multiple destinations concurrently — fan out the same RDSR to dose-tracking + EMR + ACR DIR if the institution wants overlapping coverage
Audit retention aligned with dose-monitoring regulatory expectations
What SR Engine does NOT do
Honest scope boundary. SR Engine is an SR-routing and format-translation product. The analytical / monitoring features that some customers expect from dose-tracking platforms live elsewhere.
Does NOT track dose abnormalities (alerts when a study exceeds dose thresholds, etc.) — that's Bayer Radimetrics / Sectra DoseTrack / Epic native dose tracking
Does NOT do longitudinal dose tracking per patient (cumulative dose, age-adjusted exposure) — same boundary
Does NOT generate dose-tracking reports for the QI / radiation-safety team — same boundary
Does NOT do CDS (clinical decision support) on SR content
Does route, transform, translate, and submit — those are SR Engine's job
AI-finding routing
AI vendors increasingly deliver findings as SR objects (rather than HL7 or proprietary formats). SR Engine receives, transforms, and routes those findings to the right downstream systems.
AI-vendor SR ingest (per-vendor template recognition)
PHI-shielded round-trip via keyed-hash pseudonymization (AI vendor sees only anon-IDs)
Findings re-identified on the way back and routed to PACS / EMR
Multi-vendor concurrent flows (same study, multiple AI vendors)
SR-to-HL7 / SR-to-FHIR translation for downstream systems that don't consume SR directly
Operations & administration
Same admin surface and platform features as the rest of the router-pattern family.
Web administration UI with shared sidebar shell
Role-based access control with full permission catalog
Real-time dashboard — SR ingest rate, per-template throughput, queue depth
Comprehensive audit log (21 CFR Part 11-aligned)
Webhook outputs for SIEM integration
Synthology QMS integration for change control on transformation rules
Architecture
SR producers in, SR consumers out.
A specialized variant of the router-pattern shape — narrower
than the general Router, deeper on SR-specific logic. The
producer / consumer landscape for SR is its own ecosystem
of modalities, AI vendors, dose-tracking platforms, and EMR
targets.
Use cases
Five SR-specific deployment patterns.
Most SR Engine deployments fit one or two of these primary
patterns. Multi-pattern deployments are common in larger
institutions where dose tracking, CAD, and AI findings all
flow through the same SR layer.
Pattern 1
Radiation-dose routing — the institution chooses the destination
Modalities emit Radiation Dose SR (RDSR). What the institution does with it depends on what other systems they have. SR Engine routes to whichever destination(s) match the institutional requirements — without forcing a particular dose-tracking platform.
Flow
1Modality emits RDSR after each study
2SR Engine receives, recognizes the dose-SR template, applies per-modality normalization for vendor quirks
3Route depends on institutional setup. Option A: forward to dose-tracking platform (Bayer Radimetrics, Sectra DoseTrack, Siemens teamplay) for full abnormality monitoring + longitudinal tracking
4Option B: forward to Epic Radiant — Epic's Radiology module has integrated dose tracking with ACR DIR submission, automatic RDSR capture, and XR-29 / XR-25 Dose Check alerts. For Epic Radiant institutions, no separate dose-tracking platform is needed
5Option C: direct encrypted submission to ACR Dose Index Registry — for institutions whose requirement is only the registry submission
6Option D: any combination of A / B / C concurrently
7Audit log retains the full chain for regulatory inspection
Pattern 2
SR-to-HL7 / SR-to-FHIR translation for EMR integration
Most EMRs (Epic, Cerner, others) want HL7 ORU or FHIR DiagnosticReport — not raw DICOM SR. SR Engine translates SR content into the format the consumer expects, so producers (modalities, AI vendors, reporting platforms) can keep emitting SR and the EMR side gets what it expects without the producer having to know which downstream wants what.
Flow
1Producer (modality / AI vendor / reporting platform) emits DICOM SR
2SR Engine receives, recognizes the template
3For DICOM-SR-aware consumers (PACS, dose-tracking platforms): forward SR as-is
4For HL7-only consumers (most EMR-side integrations): SR Engine transforms SR content → HL7 ORU and emits HL7
5For FHIR-aware consumers: SR Engine transforms SR content → FHIR DiagnosticReport / Observation
6Same SR fans out concurrently in different formats to different consumers — producer never has to know
Pattern 3
CAD findings distribution
Mammography CAD, Chest CAD, and other CAD applications produce SR-formatted findings. SR Engine routes those findings to PACS for diagnostic review, to reporting / dictation platforms as template inputs, to EMR via SR-to-HL7 translation where applicable, and to dose-tracking systems where dose-related CAD findings need to be visible.
Flow
1CAD application emits findings as SR (Mammography CAD SR, Chest CAD SR, etc.)
2SR Engine recognizes the CAD SR template
3Per-template routing rules forward to the right downstream destinations
4PACS receives SR for inline radiologist review
5Voice-recognition / dictation platform receives SR as template input for the radiologist's draft report
6EMR receives findings via SR-to-HL7 ORU translation
7Audit captures every CAD finding's routing path
Pattern 4
AI-finding ingestion (with PHI shielding)
AI vendors produce findings as SR. For vendors who don't take BAA scope, SR Engine handles the PHI-shielded round-trip via keyed-hash pseudonymization — same family-wide mechanism as the Router.
Flow
1Outbound: SR Engine de-identifies the source study's SR via keyed-hash pseudonymization, sends to AI vendor
2AI vendor analyzes, returns findings as SR tagged with the same anon-IDs
3SR Engine re-identifies the findings using the retained mapping
4Findings forwarded to PACS / EMR / reporting platform linked to the correct patient
5AI vendor never received PHI; clinical workflow benefits from the AI findings
Pattern 5
Modality measurement-report distribution
Ultrasound, echo, and other modalities produce SR-formatted measurement reports (TID 5200 cardiac echo, TID 5310 pediatric echo, etc.). SR Engine routes them to PACS, reporting platforms, and CVIS / cardiology PACS where applicable.
Flow
1Modality emits measurement SR alongside the imaging study
2SR Engine recognizes the measurement-SR template
3Routes to: PACS for image-context viewing · reporting platform for template population · CVIS for cardiology workflow
4Per-destination shape adjustments (each consumer wants SR slightly different)
5Confirmation ACK back to modality where required
Pattern 6
Comprehensive-SR final-report distribution
Modern reporting platforms emit the final radiologist report as a Comprehensive SR object. SR Engine routes the final report to all downstream consumers — PACS, EMR (via SR-to-HL7 translation), referring physician portals, image-sharing platforms.
Flow
1Reporting platform emits Comprehensive SR after report sign-off
2SR Engine routes to PACS for inline report retrieval
3SR-to-HL7 ORU translation for EMR systems that don't consume SR directly
4SR-to-FHIR DiagnosticReport translation for FHIR-aware downstream
5Optional fan-out to referring physician portal and image-sharing platform
6Audit captures every report's distribution path
Integration points
SR producers, SR consumers, family orchestration.
SR producers (sources)
Modalities emitting RDSR (CT, fluoroscopy, IR, angiography, mammography)
Modalities emitting measurement SR (ultrasound, echo, cardiology)
Voice-recognition / dictation platforms — including the established platforms (Microsoft Nuance PowerScribe, Jacobian M*Modal Fluency for Imaging) and the emerging AI-driven dictation systems (DAX Copilot, Abridge, Nabla, Suki, Heidi, and others entering the market)
AI vendors emitting findings as SR
Reporting platforms emitting Comprehensive SR for final reports
SR consumers (destinations)
PACS / VNA — inline SR retrieval alongside images
Voice-recognition / dictation platforms — established platforms (Microsoft Nuance PowerScribe, Jacobian M*Modal Fluency for Imaging) AND the rapidly-growing AI-driven dictation field (DAX Copilot, Abridge, Nabla, Suki, Heidi, etc.) — SR forwarded as template input for the radiologist's draft report
EMR / EHR — Epic Radiant has integrated dose tracking (ACR DIR submission, automatic RDSR capture, XR-29 / XR-25 Dose Check alerts) — for Epic Radiant institutions this satisfies ACR / state requirements without a separate dose-tracking platform. Cerner via SR-to-HL7 ORU translation
ACR Dose Index Registry — direct encrypted submission supported (no dose-tracking platform required if the only requirement is the registry submission)
CVIS / cardiology PACS for measurement reports
FHIR-aware systems via SR-to-FHIR DiagnosticReport / Observation translation
XyDromatics Router — Router can forward SR to SR Engine for specialized handling
XyDromatics VNA family — SR Engine writes to VNA Repository / Clinical / Clinical Research
XyDromatics Migration Engine — bulk SR migration during PACS replacement
Shared site encryption key for keyed-hash pseudonymization (consistent anon-IDs across products)
Identity, audit, observability
OIDC / SAML for the web admin UI
Active Directory / Azure AD federation
Per-AE-Title audit trail entries
Hash-chain audit log
Prometheus metrics + structured JSON logs
System requirements & sizing
Sized to SR object volume.
SR objects are small (typically less than 100 KB each) but
high-volume in dose-tracking and AI-finding deployments.
Sizing reflects SR object count and consumer fan-out, not
full study volume.
Tier
SR volume
CPU
RAM
Storage
Typical site
Small (single facility)
< 1,000 SR / day
2 vCPU
4 GB
100 GB (queue + audit)
Imaging center, single-modality dose-tracking deployment.
Medium (community hospital)
1,000 – 10,000 SR / day
4 vCPU
8 GB
500 GB
Single hospital with full modality mix, CAD deployment, dose tracking.
Large (academic / multi-site)
10,000 – 50,000 SR / day
8 vCPU (HA pair)
16 GB per node
1 TB per node
Academic medical center, multi-hospital, multiple AI vendors.
Enterprise (national / IDN)
> 50,000 SR / day
16 vCPU (HA cluster)
32 GB per node
2 TB per node
Large IDN with enterprise dose tracking + multi-vendor AI.
Licensing
Three packages, à la carte add-ons.
Core handles ingest + basic forwarding. Transform adds
template-aware migration and dose-SR specialization.
Enterprise unlocks AI-vendor PHI shielding, ACR DIR
submission, and HA pairing.
SR Engine Core
The base package — SR ingest, basic forwarding, audit log. Suited for single-purpose deployments.
C-STORE SCP / STOW-RS for SR
Multi-destination forwarding (up to 5 destinations)
Standard SR template recognition
Audit log + dashboard
Web admin UI + RBAC
SR Engine + Transform
Adds template-aware transformation, dose-SR specialization, and CAD-finding routing rules.
Radiation-dose SR specialization with vendor-specific routing
CAD-finding routing rules
SR-to-HL7 ORU translation
Up to 25 destinations
SR Engine Enterprise
Everything, plus AI-vendor PHI-shielded round-trip, ACR DIR submission, and HA pairing.
Everything in Core + Transform
Keyed-hash pseudonymization for AI-vendor PHI shielding
Built-in honest-broker capability for re-identification
ACR Dose Index Registry submission
SR-to-FHIR DiagnosticReport translation
Active-active HA pair
Unlimited destinations
Premium support tier eligible
Per-deployment licensing, not per-SR-object. Multi-product
bundles (SR Engine + Router + VNA family) get bundle pricing
and shared site-key provisioning.
Documentation
The controlled-document set.
Every SR Engine deployment ships with the documents below, all
managed under SynthQMS document control. The DICOM Conformance
Statement is publicly available; the rest are shared under mutual NDA.
Document
Title
Notes
DOC-2026-061
Hardware & Software Requirements (SR Engine)
Sizing, OS support, dependencies
DOC-2026-077
DICOM Conformance Statement (SR Engine)
Public — full SOP class + transfer-syntax matrix
DOC-2026-067
EULA Annex (SR Engine)
Annex to the Synthology Master EULA (DOC-2026-063)
DOC-2026-079
Penetration Test Results (SR Engine)
Available under NDA
DOC-2026-092
QA Runner Manual (SR Engine)
Internal QA + customer-validated test runs
DOC-2026-083
User Guide (SR Engine)
Operator + administrator workflows
DOC-2026-072
Installation Guide (SR Engine)
Includes MSI deploy + Linux tarball install
Frequently asked
The questions prospects ask.
Why is this a separate product from the Router?
Two reasons. First: footprint -- many customers (single-purpose dose-tracking deployments, CAD-finding distributors, AI-finding aggregators) need only the SR-specific slice of router-pattern capabilities. Deploying the full Router for them is overkill on cost and on operational surface. Second: specialization -- SR Engine is template-aware in ways that the general Router is not. Treating an RDSR or a Mammography CAD SR as opaque DICOM works, but losing the template awareness loses cross-vendor normalization quality and dose-tracking conformance. SR Engine optimizes for the SR workflow specifically.
Can we run SR Engine alongside the Router?
Yes -- common pattern. Router handles general DICOM routing; SR Engine handles the SR slice with specialized transformation. Router can forward SR objects to SR Engine for specialized handling, then SR Engine fans out to dose-tracking, CAD, reporting, etc. A customer running both gets the platform consistency benefit (shared admin shell, shared audit, shared site encryption key for the keyed-hash pseudonymization).
Do I need a dose-tracking platform like Bayer Radimetrics to use SR Engine?
Often no. SR Engine routes the RDSR; what destination satisfies your requirements depends on what your institution has. Three common shapes: (a) Epic Radiant institutions: Epic's Radiology module has integrated dose tracking, ACR DIR submission, automatic RDSR capture, and XR-29 / XR-25 Dose Check alerts. SR Engine forwards RDSR to Epic Radiant and Epic handles ACR compliance + abnormality alerting -- no separate dose-tracking platform required. (b) Non-Epic-Radiant institutions wanting full dose-abnormality monitoring + longitudinal tracking + QI reports -- you need a dose-tracking platform (Bayer Radimetrics, Sectra DoseTrack, Siemens teamplay), and SR Engine forwards to it. (c) Institutions whose only requirement is ACR Dose Index Registry submission -- SR Engine submits encrypted to the registry directly, no separate platform required. SR Engine's value is the routing flexibility; the analytical / monitoring features themselves live in Epic Radiant or in dedicated dose-tracking platforms, not in SR Engine.
Can SR Engine transform SR into HL7 messages?
Yes. For consumers that want HL7 ORU rather than DICOM SR (common for EMR-side integration), SR Engine transforms the SR content into HL7 ORU and emits HL7 messages. Same for FHIR -- SR Engine produces FHIR DiagnosticReport / Observation for FHIR-aware downstreams. The producer doesn't have to know or care which destination expects which format; SR Engine handles the per-destination shape.
Can SR Engine send SR to voice-recognition / dictation platforms?
Yes. Voice-recognition / dictation platforms take SR content as template inputs to populate the radiologist's draft report. This includes the established platforms (Microsoft Nuance PowerScribe, Jacobian M*Modal Fluency for Imaging) and the rapidly-expanding field of AI-driven dictation systems (DAX Copilot, Abridge, Nabla, Suki, Heidi, and a steady stream of new entrants). SR Engine routes to whichever platform(s) the institution uses, alongside the other consumer destinations. A single RDSR or measurement SR can land in the dictation platform (template input) AND the dose-tracking platform (compliance) AND the PACS (image-context retrieval) simultaneously. The dictation-platform landscape is moving fast; SR Engine's job is to keep routing to whatever the institution picks without constraining the choice.
What does SR Engine NOT do?
SR Engine is a routing and format-translation engine. It does NOT track dose abnormalities, does NOT do longitudinal dose tracking per patient, does NOT generate QI / radiation-safety reports, does NOT do clinical decision support on SR content. Those are the analytical / monitoring features that dose-tracking platforms (Bayer Radimetrics, Sectra DoseTrack, Siemens teamplay) or EMR-native dose-tracking (Epic) provide. SR Engine's job is making sure the right SR content gets to the right destination in the right format.
How does SR Engine handle older SR templates from legacy modalities?
Template-version migration is a core capability. Older modality firmware emits older versions of TID 1500, TID 5200, and other templates; current consumers expect newer versions. SR Engine recognizes the source template version and migrates forward to the consumer-expected version, preserving the measurement / finding semantics. The transformation is audit-trailed.
Can SR Engine produce SR for systems that don't natively emit it?
No -- SR Engine is an ingest / transform / forward engine, not an SR-creation tool. If your reporting platform produces free-text reports rather than SR, the place to address that is the reporting platform itself or the integration layer between reporting and PACS. SR Engine handles the routing and transformation of SR objects that already exist somewhere in the workflow.
How does the AI-vendor PHI-shielding workflow work here?
Same family-wide mechanism as the Router and the VNA siblings. SR Engine uses keyed-hash pseudonymization on the way out (to the AI vendor), retains the source ↔ anon-ID mapping, and re-identifies AI-returned findings on the way back. AI vendor sees only anon-IDs; findings get linked to the right patient. A multi-product customer's site encryption key is shared, so anon-IDs from SR Engine are consistent with anon-IDs from the Router or VNA Migration in the same deployment.
What's the regulatory classification?
Non-device software under FD&C Act §520(o)(1)(D), the 21st Century Cures Act §3060 carve-out for software that transfers, stores, converts, or displays device data without interpreting or analyzing it. No FDA Device Listing required. Same framework as Router, VNA Repository, and VNA Migration -- SR Engine handles storage and forwarding of SR objects without modifying clinical interpretation. Tag transformations are constrained to non-interpretive scope (template migration, anonymization, normalization).
Is SR Engine cross-platform?
Yes -- Windows and Linux, same as the rest of the router-pattern family. MSI installer for Windows, self-contained Linux tarball for Linux. Both supported equally.
Talk through your SR workflow.
Dose tracking? CAD findings? AI-finding distribution?
Comprehensive-SR final reports? Tell us the SR-specific
workflow you’re building and we’ll come back
within one business day with a proposed architecture.